An agent delegation is a structured transfer of authority that allows one agent to act, under a bounded and semantically defined scope, on behalf of another.1 visionGraph places it under Agent Relationship and Authorisation.

The core module defines the relation with delegatesTo and delegatedBy, both ranging over agents.2 The delegation module adds a delegationId, a note, and a delegationScope whose value is a Delegation Scope. It also links the delegation to the actions it permits and to trace records. The threat-model module describes the same relation as an assignment of authority or responsibility between agents, with delegatedTo and delegatedFrom.3

Delegations can be strung together in a Delegation Chain. A task handed on in this way is a Task Delegation, and an Agent Trace Event can record the hand-over.

The term is unrelated to proof-of-stake Delegation; the two contrast, and the agent sense takes this title.

Footnotes

  1. Semantic Agent Communication Community Group, “Delegation Ontology” (Agent Ontology, CC BY 4.0). https://github.com/s-agent-comm/agent-ontology/blob/750e06d276dd568313a6c03c4e78203887d3fa4b/ontologies/delegation.ttl ↩

  2. Semantic Agent Communication Community Group, “Core Agent Ontology” (Agent Ontology, CC BY 4.0). https://github.com/s-agent-comm/agent-ontology/blob/750e06d276dd568313a6c03c4e78203887d3fa4b/ontologies/core.ttl ↩

  3. Semantic Agent Communication Community Group, “Threat Model Ontology” (Agent Ontology, CC BY 4.0). https://github.com/s-agent-comm/agent-ontology/blob/750e06d276dd568313a6c03c4e78203887d3fa4b/ontologies/threat-model.ttl ↩